Concepts and methods used in correcting a digital map base with scene analysis are presented is summary form. This study is concerned with one special aspect of computer vision: How can knowledge in the form of a digital map serve in automatic image interpretation, and, on the other hand, how can interpretation results be used to change or update the map? In a wider sense, map may mean any graphic representation of a scene that is imaged. Here, in particular, this document deals with maps in the cartographic sense, and with images from airborne photographic systems. One of the obvious applications is the correction or densification of a map data base using time series of aerial surveying imagery. The aim of this study therefore is the design of a strategy to evaluate the usefulness of image-map correspondence to aid the interpretation of digital aerial photography. This is the first step to be taken towards a photo-interpretation expert system, which shall be called henceforth name PHIX.